While I wouldn't dream of standing up k8s on a bare metal cluster without a devops team, I set up managed k8s using EKS several years ago for a client and... it just chugs along, self-healing, with essentially zero maintenance.
For my own projects I use a managed Northflank cluster on my own AWS account and likewise... just a fantastic experience. Everything that Heroku could and should have been. Yes the cluster is a bit pricey to stand up both in terms of EC2 compute and management layer costs, but once it's there, it's there. And the costs scale much more nicely than shoving side projects onto Heroku.
At this stage I consider managed k8s my default go-to unless it's something so lightweight I just want to push it to Vercel and forget about it.
Happily using Cloud66 but a strong recommendation for Northflank here. I already had a running cluster there and migrating a very complex Heroku app over only took 20 minutes today. (Would have been 5 if I hadn't screwed up .gitignore...)
It struggled with tasks I asked for (e.g. download the March and April invoices for my GitHub org "myorg") -- it got errors parsing the DOM and eventually gave up. I recommend taking a look at the browser-use approach and specifically their buildDOMTree.js script. Their strategy for turning the DOM into an LLM parsable list of interactive elements, and visually tagging them for vision models, is unreasonably effective. I don't know if they were the first to come up with it, but it's genius and extracting it for my browser-using agents has hugely increased their effectiveness.
I developed a SaaS business to fund a nonprofit foundation, which is a little different from your situation.
The key thing was to keep the SaaS-y bit as boring as possible, which meant a corporation. This is in Europe but the equivalent would be Delaware C corp.
Shares in the corporation were then given to the wrapping organization (in my case the foundation, but this could be more-or-less any legal structure that can have assets). Downside is two sets of accounts, upsides are that M&A gets a ton easier later on, and taking on employees is simple and not colored by the legal quirks of the parent organization. The potential complexity of SaaS accounting (revenue recognition, R&D credits, etc) is also kept inside a simple, normal corporation which every CPA is super-familiar with, so you're not consulting niche experts every time something new comes up.
I advise a quick consult with a tax lawyer before doing anything, because it's easy to say you'll deal with this later but some changes have unforeseen implications if not done at the outset. (I punted on some of the setup for a year while I focused on finding product/market fit, and that turned out to be a mistake that the lawyers had to fix at some cost. A year more and it might have become unfixable.)
I agree that indigenous knowledge is rightly the property of indigenous peoples themselves, but that wasn't my point. My comment referred specifically to the products of evolution: that's the genetic resources, not the knowledge of them.
Doesn't the concept of biopiracy imply that the products of millions of years of evolution belong to modern nation states?
What if, instead, we either observe that they simply exist, or imbue them with their own rights, or consider them the birthright of all humanity?
We've seen that attempts to consider them as intellectual property for the purposes of ensuring their conservation have failed. We can also see that it hasn't driven the anticipated licensing revenue flows to poorer nations. But which approach would lead to greater good for humankind over a long time horizon?
Probably true - but having a great many soldiers in an area where there are an extremely isolated people such as the "uncontacted" Yanomami also poses its own risks. If they're focusing on the most sensitive areas with highly trained personnel and making use of the impressive knowledge and expertise in Brazil's Indian affairs department, FUNAI, much could be achieved. The situation in French Guiana is quite different, with perhaps only small numbers of Wayãpi living in isolation.
Brazil is rolling back the free-for-all that was established under Bolsonaro, but if they're handling it delicately, that's likely good news.
Another point, and I'd love to be corrected here, is that with a container of water you're going to get a ton of convection currents leading to a much sharper heat gradient at the edges, resulting in significant heat losses with the same amount of insulation.
At a guess, and I confess I'm not capable of running the numbers, this offsets the much higher temperature delta of sand.
For my own projects I use a managed Northflank cluster on my own AWS account and likewise... just a fantastic experience. Everything that Heroku could and should have been. Yes the cluster is a bit pricey to stand up both in terms of EC2 compute and management layer costs, but once it's there, it's there. And the costs scale much more nicely than shoving side projects onto Heroku.
At this stage I consider managed k8s my default go-to unless it's something so lightweight I just want to push it to Vercel and forget about it.